Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Intersection vehicle positioning method and system based on multiple view angles and re-identification

A vehicle positioning and re-identification technology, applied in the field of vehicle positioning, can solve the problem of high cost of multi-beam lidar, and achieve the effect of helping promotion and avoiding costs

Active Publication Date: 2022-07-22
TSINGHUA UNIV
View PDF11 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] 1. Multi-beam lidar is expensive;
[0008] 2. The vehicle-mounted lidar can only obtain environmental information within a certain range of the vehicle itself from the perspective of the vehicle;
[0010] 4. The vehicle-mounted automatic driving system needs to process huge data information in a limited storage space

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Intersection vehicle positioning method and system based on multiple view angles and re-identification
  • Intersection vehicle positioning method and system based on multiple view angles and re-identification
  • Intersection vehicle positioning method and system based on multiple view angles and re-identification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0061] like figure 1 As shown, Embodiment 1 of the present invention proposes a method for locating vehicles at intersections based on multi-view and re-identification, figure 2 It is the overall model structure diagram, and the specific implementation steps are as follows:

[0062] Step 1) The roadside equipment preprocesses the raw data of the multi-view camera, including operations such as whitening and denoising;

[0063] Step 2) The roadside equipment performs feature extraction on multi-view images, including: using ResNet50 as the backbone network, and introducing the design of "bottleneck design", in order to reduce the calculation time, two 3 × 3 convolutions are replaced by 1 ×1+3×3+1×1 convolutional layers; in order to obtain a lightweight feature extraction model, the number of channels is reduced from 256 to 64 through the first 1×1 convolutional layer, and then 1×1 1 for convolution recovery.

[0064] Step 3) Feature projection and feature fusion: specificall...

Embodiment 2

[0080] Embodiment 2 of the present invention proposes an intersection vehicle positioning system based on multi-view and re-identification. The system is set in a drive test device and includes a preprocessing module, a bird's-eye view generation module, a search domain generation module, a re-identification module and a positioning module , a positioning sending module and a vehicle detection model, where,

[0081] The preprocessing module is used to preprocess the original image of the intersection collected by the multi-view camera;

[0082] The bird's-eye view generation module is used to sequentially perform feature extraction, perspective transformation and aggregation on the preprocessed view images to generate a bird's-eye view of the distribution of vehicles at the intersection;

[0083] The search domain generation module is used to input the original image of the intersection into the pre-established and trained vehicle detection model to obtain the search domain of...

Embodiment 3

[0088] Embodiment 3 of the present invention may further provide a computer device, comprising: at least one processor, a memory, at least one network interface and a user interface. The various components in the device are coupled together by a bus system. It can be understood that the bus system is used to realize the connection communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus and a status signal bus.

[0089] Among other things, the user interface may include a display, a keyboard, or a pointing device (eg, a mouse, a trackball, a touch pad or a touch screen, etc.).

[0090] It can be understood that the memory in the embodiments disclosed in the present application may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Wherein, the non-volatile memory may be Read-Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an intersection vehicle positioning method and system based on multi-view and re-identification, and the method is realized based on a multi-view camera deployed on roadside equipment, and the method comprises the steps: carrying out the preprocessing of an original image of an intersection collected by the multi-view camera; sequentially performing feature extraction, perspective transformation and aggregation on the preprocessed visual angle image to generate a distribution aerial view of vehicles at the intersection; inputting the original image of the intersection into a pre-established and trained vehicle detection model to obtain a search domain of vehicles at the intersection; receiving a look-around picture and id information sent by a vehicle at an intersection, and re-identifying the vehicle in the search domain to obtain a picture of the vehicle at a certain visual angle; performing feature extraction and perspective transformation on a certain visual angle image obtained by re-identification, and projecting the image to the aerial view to obtain position information of the vehicle in the aerial view; and sending the position information and the id information of the vehicle to the vehicle so as to complete positioning.

Description

technical field [0001] The invention belongs to the field of vehicle positioning, in particular to a method for locating vehicles at intersections using multiple viewing angles and re-recognition, in particular to a method and system for locating vehicles at intersections based on multiple viewing angles and re-identification. Background technique [0002] Nowadays, artificial intelligence technology has been widely used in the field of automatic driving, and the positioning accuracy has greatly increased, which has greatly improved the safety of the automatic driving process. In the field of autonomous driving, vehicle detection and localization are essential steps, which directly affect the safety of autonomous driving. Most of the methods that can extract precise positioning information are based on the point cloud information collected by lidar. However, although lidar can achieve high positioning accuracy, it is expensive. [0003] Multi-view image information, as a p...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06V20/40G06K9/62G06N3/04G06N3/08G06V20/52G06V10/74
CPCG06N3/08G06N3/045G06F18/22Y02T10/40
Inventor 李骏张新钰李志伟黄朝胜金宪伟周沫吴新刚李金伦
Owner TSINGHUA UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products